NickNaraghi

This makes me surprisingly excited for the iPhone Duo - I didn't like it at first, but seeing it through this lens, seems like Apple is competing for the "AI-native device" of the future.

There was a really cool project a few years ago from Area Tech that explored this - generative shields (with heraldry on them). https://shields.build/

Seems like this would have strong implications for distillation and/or smaller types of transformers!

You didn't say please or thank you.

Trailer link: https://www.youtube.com/watch?v=dPhJLtzKk68

This[0] continues to be one of the most useful articles I’ve ever read.

[0]: https://www.slatestarcodexabridged.com/Meditations-On-Moloch

Feels like a major unlock with 1) how good AI has gotten, 2) how fast it can output meaningful work. Excited that there's a whole new design space opening up! This feels like one of the first truly new things I have seen.

Rediscovered here https://news.ycombinator.com/item?id=439065

No don’t go, it sucks! It’s dusty and hot!

I wonder if we’ll get something like CDLMs for automated harness engineering, sort of piloting the LLM underneath.

Will check it out, thanks!

Thanks for the thoughtful write-up. I’m currently working on a journal article about optimizing organizations from first principles, starting with optimized delegation. Would love your thoughts on the paper if you’re willing to take a look.

These are two public previews that we are building off of for the journal:

1. https://paragraph.com/@hq.spengrah/an-actionable-model-of-tr...

2. https://x.com/hatsprotocol/status/1986458735095517526

> Across his various startups, Peter has seen two kinds of work:

> 1. the "IQ 180" work. some mad scientist genius type comes up with some crazy solution you've never thought of.

> 2. the "token spewer" work. being ultra responsive, pushing the ball forward across dozens of different fronts.

Interesting comp to pg's Maker's Schedule, Manager's Schedule https://www.paulgraham.com/makersschedule.html

I'm curious about not only which of these roles models will fill, but also how they will empower us to be in the mode we prefer.

I feel that more people would understand Anthropic's worldview if they read more of the public source material. Sure it is a cult, but it they have been relatively transparent with their views: https://darioamodei.com/essay/machines-of-loving-grace#basic...

> Let’s say every company gets about three innovation tokens. You can spend these however you want, but the supply is fixed for a long while.

This is one of my favorite blog posts, and it can basically be encapsulated in the idea of "innovation tokens." It is one of the most useful concepts I have had as a PM / eng leader in my career. It helps actually make the the right tradeoffs, and helps even more in explaining those tradeoffs to colleague of all levels. Highly recommend.

> You built AGI on Mar 15, 2033. Ada built superintelligence on Aug 12, 2033.

Pretty fun. A little one-dimensional but a good insight into how a VC thinks about RSI. My second attempt was almost exactly the same outcome as the first, despite a different capital allocation.

These used to be 50 cents a piece at my high school library. What a joy it was to have a tiny simple way to transport essays (and pixel art and other fun things) to and from school and amongst all the school computers.

Seems like an artifact of the subagent pattern which is explicitly included in recent models.

I think one day it will relatively plateau and then these approaches will be meaningful improvements in performance. But yes, pointless for now.

I would put Reply and Report on the same line as the poster anon username, to condense even more

This is awesome! Feedback would be to increase the visual density of the posts / replies (taking from HN or Reddit) because right now there's a lot of space between each line.

Unfortunately it is a necessary compromise, because free speech is essential to a working civilization.

I think your reasoning is actually more complex than the person you are responding to, unfortunately. (As someone who as published broadly used open source software)

Really cool. What do the buttons on the right sidebar do?

Wild how much empathy I have for the LLM now. It will definitely affect how I use AI.

These are particularly interesting because they isolate the one part of time that is completely made up: the week. Days, months, years have natural markers - but weeks and particularly days of the week are a totally arbitrary coordination function.

You had two weeks to build up a queue of everything you wanted to do with it!

Gonna have to see the agent trace on that one.

Almost every correct answer is a longer string than the other multiple choice options.

Super interesting. How does regulation affect IPO prospects of OAI and ANTH? What’s the timing of each?